Inspiration

Motivation:

=> Covid-19 takes a toll on mental health; When people experience traumatic and often unexpected events, the reaction is usually a combination of fear and distress.

=> Many resources focus on detection of physical well-being, but not on mental well-being. E.g. The impact of Covid-19 is often measured by the number of fatalities, number affected, and recovered; But how do we measure the mental toll?

What do we do about this problem?

=> We can detect suicide / depression clusters; clusters of population that are most susceptible to suicidality or depression.

Why are suicide / depression clusters important to identify?

*All of these clusters lead to risk of sudden loss in population, panic, and pain:

=> Point clusters (same geographic location within the same time period)

=> Mass clusters (across different geographic locations but within same time period)

=> Echo clusters (same geographic location, at a later time period)

How do we detect clusters?

=> Certain factors increase the risks of suicide;

  • Age: older people are particularly vulnerable during this pandemic, and already suffer from high rates of loneliness, which is strongly associated with greater symptoms of depression, and mortality
  • Physical morbidities: evidence shows inadequate physical healthcare can lead to greater risk of mental illness
  • Number of deaths by Covid-19 confirmed in the area: exposure to another person’s suicide, be it acquaintances, family, or person in local viscinity - can drastically increase risk of suicide clusters.
  • Number of active cases of Covid-19 in the area: large no. of cases leads to increased anxiety and fear, and due to people feeling ‘unsafe’ about going outside, staying indoors leads to increased loneliness / isolation. These all lead to increased risk of suicide/depression.

So. How do we retrieve this data?

=> We gather data using the Google Cloud BigQuery Public Dataset: Covid-19 Open Data, using BigQuery!

=> The dataset provides access to critical information quickly, easily & is automatically retrieved.

The structure of this list is as follows: Previous dataset column name : COVID-19 Open Data column name province_state : subregion1_name country_region : country_name date : date latitude: latitude longitude : longitude location_geom : location_geometry confirmed : cumulative_confirmed deaths : cumulative_deceased recovered : cumulative_recovered active : current_cases (NOTE: Calculated as [cumulative_confirmed-cumulative_recovered-cumulative_deceased] where none of these values are NULL) fips : subregion2_code admin_2 : subregion2_name combined_key : location_key

Where does our App fit in?

By mapping out high-risk clusters using metrics provided by the automatically retrieved Covid-19 open data, we can:

  1. Reallocate scarce medical resources,
  2. Optimize the stationing of first-responders and police for injury / emergencies.

*Our application will act as a virtual ‘cluster response group’; it helps to effectively identify and respond to mental high-risk clusters.

Implementation

Web application

What it does

Effectively relays information about 'high mental-health risk clusters' to first responders and police dispatchers. Shows active alerts for areas that exceed a risk threshold.

How we built it

App retrieves real-time data from the Covid-19 Open Database on the Google Cloud Platform, and configures Firebase as a database to store and update this data, which is then processed into a 'risk level / category' to be displayed for the users.

We used the Google Maps and Places API to map out clusters of high risk populations across the US. When the user submits their region through the website, the map automatically adjusts for the submitted locality.

What we learned

Lots of frontend & backend technology, and using Firebase with Google Cloud BigQuery! Working together virtually, and communication skills.

What's next for MyRiskCluster

Follow our repo on github to see get project updates!

Share this project:

Updates